Thoughts on the strength and reliability of integrating a transmission like this? We’re mostly worried about the possibility of the gears eating each other with the added torque the FP gives. As shown the reduction is 12.75:1 but the optional gearings available for the toughbox also work here.
The 20DP, 14.5PA steel AM Toughbox gears should be more than capable of transmitting that much torque (and much more) before giving you any trouble. Several teams have used similar arrangements in the past.
My only other thought is that I’m not sure if the relatively meager weight savings of dogging the inner plate like that are worth the decrease in strength.
If the frame flexes slightly from an impact (unlikely with bumpers on, but still) what is the probability of the gearboxes failure?
And yes, that is a FP through the new AM Planetary casing (saves .2lb!). I do miss the blue anodized and red anodized casings though… maybe Andy still has some laying around. I left the inner frame visibility off so one can actually see the two stages (mostly). In the animation, the inner frame is visible but I removed visibility on one of the standoffs holding the mounting plate away from the inside plate.
Remember that if you use the 14t CIM gear from AM you will not be able to pull the motor and pinion assembly through the CIM pilot hole without taking it apart.
You could make the center CIM mount hole larger than the pinion OD(.9" or so), then tighten your tolerance on the bolt holes to give you some alignment…this way you can disassemble the motors without having to remove the pinions.
Or just use the 12t pinion that is used on SuperShifter.
Thanks for the input ajlapp, I think we’ll take the advice to allow us to pull the motors out without banging at em with a flathead first.
Here is a picture of the mounting plate. Things circled in red are motor mounts. Things circled in purple are where the spacers go. The bit circled in lime-green are the two reductions for the gearbox.
We ended up keeping the mounting holes the same dimensions as the TB so if things go south we can just bolt a pre-assembled toughbox on and keep on rolling.
As long as both the frame and the transmission plate aren’t too flexible and have a small amount of “loose” tolerance, there shouldn’t be any problems. Team 228 had a similar design last year, taking the guts out of the Toughbox (except with a 12t CIM pinion gear instead of the 14t) and building it into the chassis with the output shaft of the gearbox direct driving the rear wheels of the robot. Our main competition robot used a single 0.125" CNC sheet metal plate without standoffs for the transmission, and the practice robot used a 0.1875" CNC-milled aluminum plate with standoffs. Both performed equally well, with the competition robot running an entire season (six competitions) and countless hours on the practice robot without any problems. Here’s some photos:
I would highly suggest against pocketing your motor mounting plate like that. Not for strength concerns, but for cost and time. The weight savings simply aren’t worth the headache.
That being said, Nice design! Ought to make everything nice and integrated. If you can figure out a system to make each side of the drive system modular (IE GRT in 2007), it would add to the system’s robustness and quick-changeability during competition.
I’m still waiting for a team that can pull of Nascar speed pit stops for battery and drive changes. It’ll be cool to see a team rush into the put, drop the entire robot base (drive+battery), latch a new one on, and rush off. You know, I think my CAD setup is calling me…
I wouldn’t say this is a universal truth. Assuming they are getting plates CNC’d anyway, it’s really not a headache at all, and not a lot of additional run time.
I know last year we took out about 1.5 lbs total from our gearbox plates, even if was half a pound, that could be a big deal.
And even if the plates aren’t being CNC’d, pocketing can be done by hand or on a manual relatively easily, and an hour of work for half of pound of weight is a no brainer for any team (there are an awful lot of hours in the season…).
Each drive assembly contains the gearbox, two sideplates bolted together with spacers, and the wheel assembly. Thusly with all assemblies added together, you get 1 drive assembly (1/2 of the bot, 2 wheels). Each drive assembly bolts to 3 1"x1" 8020 bars that go across the bot. So technically, if we had backup full drive assemblies made, we could fully change out half of our drive train by removing 3 bolts, sliding the drive assembly off the 8020, putting the new assembly on, and bolting the drive assembly back to the 8020 crossbeams. I’d reckon with a pre-made backup, it could be a 2 minute job (Or, if you wanted a super-speed run nascar style, a power drill + bot on a crate + a guy holding the assembly beside you, could be a 20 second job). I think this older picture of our drive assembly shows what i’m talking about nicely, along with the picture of our 2009 bot. We used this same 8020 crossbeam design in 2007 and 2009 and they worked excellently
Team 696 did two CIMs and an FP with AM Planetary all meshing with a single spur gear in 2005. It worked just fine. I believe it was 16 tooth on the CIMs and 14 on the AM Planetary (for the FP that year).
On the subject of tooth counts for CIM pinions, if the design allows, I would choose a 14 tooth over a 12 tooth, because it will have better meshing characteristics and less wear. I believe a 12 tooth 20DP is undercut (a 14 may be also, but a 12 even more so) which is not preferred. In either case, it would be wise to run a hardened pinion. I believe the Martin S20XX series gears are harder than the AM, or at least have some sort of surface coating that resists wear. We had problems with the as-received 4140 AM 14 tooth gears wearing, even with lubrication. In theory, for a long service life, the pinion should have no fewer than 17 teeth, and the stage ratio should be no greater than 4:1, but in FRC we often break these rules due to the short operation cycles we see here.
Has anyone tried “backyard” hardening of gears? I’d be worried about scale produced on the surface, that would have to somehow be removed.
Huh, just for kicks and giggles I asked a sponsor of our team if they could waterjet out this plate and what do you know, 12 hour turn-around from sending the DXF and getting it made. I’m so very happy right now.